© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 26485
    Prev
    Next
    INDEX
    Explanations

    former vs latter

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æ¯ĹéĤ»
    -0.09
    å¥ģ
    -0.09
    assin
    -0.08
    æĮij
    -0.08
    æŀ§
    -0.08
    åīĶ
    -0.08
    éļĬ
    -0.08
    Keith
    -0.08
    示èĮĥåĮº
    -0.08
    ering
    -0.08
    POSITIVE LOGITS
    åIJİèĢħ
    0.26
    åīįèĢħ
    0.24
     latter
    0.19
    Former
    0.13
    两èĢħ
    0.12
     former
    0.11
    äºĮèĢħ
    0.11
     Former
    0.11
     Latter
    0.10
    ìĬ¹
    0.09
    Activations Density 0.060%

    No Known Activations